PHYSICAL PARADOXES. 



pass from one to the other, and boiling 

 ceases. 



The actual experiment is easily performed 

 in either of two ways. 



A piece of ice may be dropped into a quan- 

 tity of liquid air. For the first few moments 

 the spheroidal state prevails. The ice is too 

 hot to touch the liquid, and there is only very 

 brisk vaporisation. But being a bad con- 

 ductor, the ice soon has its surface cooled suffi- 

 ciently for active boiling, which is maintained 

 by the heat slowly conducted from the centre 

 of the lump. 



Or a hollow may be cut in the top of a block 

 of ice, and the liquid air poured into it, when 

 the same series of events follows. The latter 

 method offers an opportunity for a pretty and 

 striking effect. As the nitrogen is boiled away 

 first, the oxygen left behind supports brilliant 

 combustion of a piece of wood, and the bright 

 light showing through the ice looks very fine 

 in a darkened room. 



174 



